Money Recognition for the Visually Impaired: A Case Study on Sri Lankan Banknotes
Akshaan Bandara
TL;DR
This work tackles accessible currency recognition for visually impaired users by developing an offline, real-time, smartphone-based system for Sri Lankan banknotes. It fine-tunes EfficientDet on a custom, smartphone-taken dataset and reports high validation accuracy with $AP=0.9847$ and $mAP=0.9877$, demonstrating strong detection performance. The approach delivers an end-to-end on-device pipeline, including data collection, augmentation, on-device deployment via TensorFlow Lite, and voice output through TextToSpeech, enabling independent denomination identification without internet access. While effective, the study notes limitations such as night-mode usability and minor latency, and points to future work expanding to other currencies and exploring newer mobile frameworks to increase accessibility and global applicability.
Abstract
Currency note recognition is a critical accessibility need for blind individuals, as identifying banknotes accurately can impact their independence and security in financial transactions. Several traditional and technological initiatives have been taken to date. Nevertheless, these approaches are less user-friendly and have made it more challenging for blind people to identify banknotes. This research proposes a user-friendly stand-alone system for the identification of Sri Lankan currency notes. A custom-created dataset of images of Sri Lankan currency notes was used to fine-tune an EfficientDet model. The currency note recognition model achieved 0.9847 AP on the validation dataset and performs exceptionally well in real-world scenarios. The high accuracy and the intuitive interface have enabled blind individuals to quickly and accurately identify currency denominations, ultimately encouraging accessibility and independence.